EP0177076A1 - Circuit for converting an analog television signal into digitized colour signals - Google Patents

Abstract

In a circuit arrangement for deriving digital color signals from an analog television signal with an A / D converter (1) which forms a digital television signal from the analog television signal in time with a clock signal which corresponds to an integral multiple of the line frequency of the television signal, and a demodulator (18), a first phase detector (25), a first phase arithmetic unit (37) and a signal generator (41) which are connected to one another in the manner of a digital phase locked loop, this results in interference-free demodulation of the chrominance signal in the television signal even when the frequency of the clock signal fluctuates ensures that the first phase arithmetic unit (37) in addition to a color phase signal from the first phase detector (25) continues to be supplied by a reference circuit (45, 47, 52, 55) a reference phase signal which is a measure of the frequency or phase difference between the clock signal and a frequency stable reference signal.

Description

The invention relates to a circuit arrangement for deriving digital color signals from an analog television signal with an analog-digital converter which forms a digital television signal from the analog television signal, which digital sequence signal comprises a sequence of amplitude-discrete samples of the analog television signal with a repetition frequency determined by a clock signal, which corresponds to a multiple of the line frequency of the television signal, a demodulator which forms the digital color signals by multiplying the digital television signal by at least one digital demodulation signal, a first phase detector which derives a color phase signal from certain digital color signals which is a measure of the phase positions of the digital color signals relative to the demodulation signals, a first phase arithmetic unit which emits a first control signal which consists of a sequence of values with a sequence frequency determined by the clock signal and in which the difference is two in each case he successive values are set by the color phase signal, and a signal generator which forms the demodulation signal or signals from the control signal, and a reference circuit which comprises a reference signal source, a phase detector generating a reference phase signal and a second phase arithmetic unit which comprises a second one of a sequence of Outputs existing control signal signal values, the difference between successive signal values is set by the reference phase signal, and which feeds the reference phase signal to the first phase arithmetic unit for additional adjustment of the difference between two successive values of the first control signal.

Such a circuit arrangement is known from European patent application 111 981. The circuit arrangement described there comprises a phase arithmetic unit, to which a clock signal, the frequency of which corresponds to an integral multiple of the line frequency of the television signal, is supplied and which outputs a control signal. This control signal consists of a sequence of amplitude-discrete values that have a repetition frequency that corresponds to the frequency of the clock signal. The values of the control signal address in a read-only memory (signal generator) samples of a sinusoidal and a cosine-shaped signal that are sent to an output with a repetition frequency according to the frequency of the clock signal are given. The frequency of the sine or cosine signal is equal to the frequency of the color burst signal from the television signal. The analog television signal is sampled in an analog-to-digital converter at the frequency of the clock signal and thus converted into a clock-frequency sequence of amplitude-discrete digital samples. This digital television signal is multiplied in two multipliers by the sinusoidal or cosine-shaped signal from the read-only memory. The digital color difference signals thus formed are fed to a phase detector, which generates a color phase signal. This is fed to the phase arithmetic unit for correcting the difference between two successive values of the control signal supplied to the signal generator.

The determination of the clock frequency to an integer multiple of the line frequency of the television signal has the advantage over a determination of the clock frequency to an integer multiple of the frequency of the color synchronizing signal that the temporal succession of the samples of the television signal is broken down into lines, fields and frames are immediately adjusted. This considerably simplifies the use of digital image storage arrangements and improves the image quality achieved in the process. The processing of television signals of different standards is also simplified.

In the simplest case, a phase arithmetic unit of the known type comprises a modulo adder with two inputs and an output, which is connected to an input of a register, the output of which is looped back to one of the inputs of the modulo adder. A digital signal, the value of which is also referred to as an increment, is fed to the modulo adder via the second input. A clock signal is also fed to the modulo adder. In each period of this clock signal, the values of the signals present at the inputs of the modulo adder are added to one another and the sum signal thus formed is fed to the register via the output of the modulo adder and stored there. The value stored in the register in one period of the clock signal is fed to the first input of the modulo adder and combined with the increment to form a new sum signal in the next period of the clock signal.

Such a phase arithmetic unit essentially represents an accumulation device that accumulates the increment with the frequency of the clock signal. The increment forms the difference between two successive values of the sum signal and thus of the signal output at the output of the register. When accumulating, the sum signal is thus gradually increased from the initial value at the frequency of the clock signal until a certain value is exceeded and the accumulation starts again with the residual value. The sequence of the values of the phase locked loop at the output of the The signal emitted in the register thus forms a time-discrete periodic signal with a frequency that depends on the frequency of the clock signal and the increment. This makes it possible on the one hand to set the frequency of the signal emitted by the phase arithmetic unit with the increment, but on the other hand fluctuations in the frequency of the clock signal, which can be caused in particular by fluctuations in the line frequency of the television signal, have a disadvantageous effect. Such variations typically occur in image recordings of devices, such as video recorders, supplied _F ernsehsignalen on. The sinusoidal or cosine-shaped signals, which are used as demodulation signals for demodulating the color beard signal from the television signal, give rise to fluctuations in the color difference signals, which lead to unpleasant color disturbances in the television picture. It can be seen that the television picture is much more sensitive to interference from the color disturbances caused by a certain deviation in the frequency of the clock signal than to the deviation of the line frequency corresponding to this deviation in the frequency of the clock signal.

In the known circuit arrangement, a reference circuit has therefore been added which generates a reference phase signal which is fed to the first phase arithmetic unit in order to compensate for fluctuations in the frequency of the clock signal which are caused by fluctuations in the line frequency. The reference circuit also generates the line-frequency clock signal with the aid of the second phase arithmetic unit. The reference phase signal serves as an increment of the second phase arithmetic unit, which accumulates the increment with a frequency of a reference clock signal generated by the reference signal source. In a downstream of the second phase arithmetic unit a square wave signal from a sawtooth signal forming converter, which comprises at least one D / A converter, the clock signal is formed, which is also supplied to the second phase detector via a frequency divider. The second phase detector, to which line sync pulses are also fed, generates the reference phase signal. Since the frequency of the reference signal source is much greater than the frequency of the clock signal, fast and therefore complex digital modules and also a fast D / A converter are required for the stages operating with the frequency of the reference clock signal.

The invention has for its object to provide a circuit arrangement of the type mentioned, which ensures trouble-free demodulation of the chrominance signal in the television signal with less effort and fluctuating frequency of the clock signal.

This object is achieved in a circuit arrangement of the type mentioned at the outset in that the sequence of signal values of the second control signal of the second phase arithmetic unit consists of a repetition frequency determined by the clock signal and that the second phase detector compares the second control signal with the reference signal generated by the reference signal source and therefrom forms the reference phase signal.

The reference phase signal, which is fed as an increment to the first phase arithmetic unit together with the color phase signal, is also supplied to the second phase arithmetic unit as an increment, resulting overall in an arrangement of two phase-locked loops coupled to one another. The phase locked loop consisting of a second phase arithmetic unit, reference signal source and second phase detector controls phase shifts caused by fluctuations in the frequency of the clock signal first control signal by changing the reference phase signal very quickly, whereby the influences of the fluctuating clock signal on the first phase arithmetic unit are compensated for very quickly without the phase-locked loop, to which the first phase arithmetic unit belongs, having to have a correspondingly high control speed. The specified arrangement of two digital phase locked loops coupled to one another is thus stabilized against fluctuations in the frequency of the clock signal. This stabilization is determined by the stability of the frequency of the reference signal, for which purpose the reference signal source preferably comprises an oscillating crystal, from the oscillations of which the reference signal is derived. The entire circuit arrangement is thus supplied by a clock signal generator and the reference circuit consists of digital components that are easy to implement.

In principle, in the circuit arrangement according to the invention the frequency of the clock signal can be a rational number (fractional) multiple of the line frequency of the television signal; it is preferably an integer multiple.

According to a development of the invention, the reference phase signal is additively superimposed on the color phase signal before being fed to the first phase arithmetic unit. This superposition is carried out in particular in an adder connected upstream of the first phase arithmetic unit, so that only a signal for setting the difference between two successive values of the first control signal is fed to the first phase arithmetic unit.

According to a further development of the invention, in which the clock signal is derived in a clock signal phase locked loop from a clock oscillator, the oscillation frequency of which is determined by a frequency control signal obtained from the comparison of a line synchronization signal contained in the analog television signal with the frequency-divided clock signal, the frequency control signal is at least the reference phase signal supplied to the first phase arithmetic functionally superimposed. The frequency control signal already contains information about the fluctuations in the line sync signal and thus about the upcoming fluctuations in the frequency of the clock signal, even before these are detected by the second phase detector and corrected by the second phase arithmetic unit. This information about the fluctuations in the frequency of the clock signal is then fed to the first and the second phase arithmetic unit as an increment in the sense of a rough regulation of the fluctuations in the frequency of the clock signal that occur, by means of additive superposition with the reference phase signal. Following this rough regulation, the specified, precise regulation is then carried out by the coupled phase-locked loops.

According to a further embodiment of the invention, the reference phase signal is fed to the first and the second phase arithmetic unit via a reference phase signal controller. By inserting such a reference phase signal controller into the phase locked loop, its control behavior, for example the control speed, can be influenced in the desired manner. In the present case, a reference phase signal controller with a high control speed is preferably selected.

According to a further embodiment of the invention, the reference phase signal controller comprises a proportional-integral control stage, also referred to for short as a PI controller. With such a PI controller, fast control is possible on the one hand and precise control on the other hand.

According to another development of the invention, the reference phase controller comprises a chain connection of at least two integrators, to which at least one proportional branch is connected in parallel. The arrangement of a second phase arithmetic unit, a second phase detector and a reference phase signal controller then forms a phase locked loop of at least a third degree. With such a reference phase signal controller, the control error of the phase locked loop is reduced for certain courses of changes in the frequency of the clock signal, in particular for changes that are linear over time. In particular, if the reference phase signal controller is designed with only one PI controller and a continuous, ramp-like change in the frequency of the clock signal, a constant control error can occur, which then supplies an incorrect increment to the first phase arithmetic unit during this ramp-like change. Such control errors are avoided with a third-degree phase locked loop.

According to a further embodiment of the invention, the reference phase signal is fed to an input of a distortion stage and its output signal is fed to the reference phase signal controller and the distortion stage transforms the reference phase signal in such a way that large values of the reference phase signal are increased disproportionately compared to small values according to a non-linear assignment between values at the input and at the output . The phase-locked loop then has small deviations between the Frequency and / or phase of the reference signal and the second control signal have a small control gain, ie that small deviations only lead to a small value for the reference phase signal. The phase locked loop then also has a low control speed and a small bandwidth, ie good noise suppression with precise, slow control of small control errors. Due to the disproportionate increase in the reference phase signal at the output of the distortion stage compared to the signal at its input with larger deviations in the frequencies and / or phases of the reference signal and the second control signal, however, it is achieved that the phase locked loop has a high control gain and thus a high control speed when large disturbances occur receives. It can quickly correct large phase and / or frequency deviations. The associated larger bandwidth and thus worse noise suppression of the phase-locked loop does not have a disadvantageous effect, since the loop has not settled anyway when it occurs.

According to a further development of the invention, the distortion stage comprises a memory arrangement which forms a value at the output of the distortion stage for each value of the reference phase signal and which supplies this to the reference phase signal controller. Such a memory arrangement can be designed as a read-only memory, the individual memory locations of which are addressed by the values of the signal at the input of the distortion stage and which then supplies the values stored in the addressed memory locations to the output of the distortion stage. In this way, any assignment between the values at the input and the values at the output of the distortion stage can be achieved. If necessary, several memory arrangements with different assignments can optionally be used and can optionally be switched into the phase-locked loop, for example, by switching devices.

According to another development of the invention, the distortion stage comprises a quantizing stage, the values of the reference phase signal. compares with one or more threshold values, as well as a selection circuit controlled by the quantizing stage, which converts values of the reference phase signal supplied to its input into individual ranges of values limited by the threshold value or thresholds according to a linear assignment into values which are fed from the output of the distortion stage to the reference phase signal controller. With such an arrangement, a linear allocation of the values at the input and at the output of the distortion stage is obtained. A small control gain can then be set for small phase and frequency deviations and a larger control gain for large phase and frequency deviations. Compared to an addressable memory arrangement, a reduction in the circuit complexity can be achieved in particular with a small number of threshold values.

According to another development of the invention, the reference signal source emits a rectangular signal which is sampled with a register controlled by the clock signal. This register is preferably designed as a 1-bit register which stores the two signal values of the rectangular signal.

According to another embodiment of the invention, the reference signal source comprises an oscillator which emits a signal at a frequency which corresponds to an integer multiple of the frequency of the reference signal, and a conversion circuit which converts the signal of the oscillator into converts an amplitude-discrete signal that is periodic with the frequency of the reference signal. Such a conversion circuit is designed, for example, as a counter or as an analog-to-digital converter. Values of the second control signal and of the reference signal are then fed to the second phase detector with a repetition frequency corresponding to the frequency of the clock signal. If, in particular, the frequency of the signal emitted by the oscillator is significantly higher than that of the reference signal, the conversion circuit carries out an exact quantization of the reference signal, as a result of which an exact phase comparison is possible in the second phase detector. In the phase-locked loop in which the second phase detector is arranged, only a small quantization noise occurs, which is caused by the amplitude-discrete reference signal. As described above, the phase-locked loop can then have a large bandwidth, ie a high control speed, without substantial noise occurring in the loop.

According to another embodiment of the invention, the frequency of the reference signal is at least almost equal to the frequency of the color burst signal of the television signal. Such a choice of the frequency of the reference signal is expedient since the first control signal emitted by the first phase arithmetic unit has the frequency of the color synchronizing signal. If the frequency of the reference signal at least almost coincides with that of the color synchronizing signal, a particularly simple circuit arrangement is obtained overall.

• _Fig. 1 ein Blockschaltbild eines Ausführungsbeispiels,
• Fig. 2 ein etwas detaillierteres Blockschaltbild eines Beispiels für ein Phasenrechenwerk aus der Schaltungsanordnung nach Fig. 1,
• Fig. 3 ein etwas detaillierteres Blockschaltbild einer Ausführungsform für eine Proportional-Integral-Regelstufe in der Schaltungsanordnung nach Fig. 1,
• Fig. 4 ein Blockschaltbild einer weiteren Ausführungsform,
• Fig. 5 ein etwas detaillierteres Blockschaltbild eines Ausführungsbeispiels für eine Referenzsignalquelle sowie einen zweiten Phasendetektor aus der Schaltungsanordnung nach Fig. 4,
• Fig. 6 eine Abwandlung eines Teils der Schaltungsanordnung nach Fig. 5,
• Fig. 7 ein etwas detaillierteres Blockschaltbild einer Verzerrungsstufe aus der Schaltungsanordnung nach Fig. 4.

Some embodiments of the invention are shown in the drawing and are described in more detail below. Show it

• _F ig. 1 shows a block diagram of an exemplary embodiment,
• FIG. 2 shows a somewhat more detailed block diagram of an example of a phase arithmetic unit from the circuit arrangement according to FIG. 1,
• 3 shows a somewhat more detailed block diagram of an embodiment for a proportional-integral control stage in the circuit arrangement according to FIG. 1,
• 4 shows a block diagram of a further embodiment,
• 5 shows a somewhat more detailed block diagram of an exemplary embodiment for a reference signal source and a second phase detector from the circuit arrangement according to FIG. 4,
• 6 shows a modification of part of the circuit arrangement according to FIG. 5,
• 7 shows a somewhat more detailed block diagram of a distortion stage from the circuit arrangement according to FIG. 4.

In the circuit arrangement according to FIG. 1, 1 denotes an analog-to-digital converter, to which an analog television signal is supplied at an input 2, which contains an analog luminance signal, an analog color tone signal and vertical and horizontal synchronizing pulses. A periodic clock signal is fed to the analog-to-digital converter 1 at a clock input 3. In each period of this clock signal in the analog-digital converter 1, an amplitude-discrete sample value is derived from the analog television signal at the input 2 and is preferably output in parallel form at an output 4.

The digital television signal is fed on the one hand to an input 5 of a filter stage 6, which in a first step obtains a digital luminance signal by suppressing the hue-tone signal and outputs it to a luminance signal output 7 and in a second step Step by low-pass filtering wins a signal that essentially still contains the sync pulses and low-frequency components of the luminance signal and that is output at a sync pulse output 8. From there, this signal reaches the input 9 of a synchronous isolating stage 10, which separates the vertical synchronizing pulses therefrom and outputs it to a vertical synchronizing output 11. The horizontal synchronizing pulses are fed from the synchronizing isolating stage 10 via a horizontal synchronizing output 12 to a synchronizing input 13 of a clock signal generator 14. The clock signal generator 14 outputs the clock signal at a clock signal output 15, which is fed to the clock input 3 of the analog-digital converter 1. The clock signal generator 14 supplies 16 line synchronizing pulses, which are derived from the horizontal synchronizing pulses, at a line synchronizing output.

The clock signal generated by the clock signal generator 14 is synchronized by the horizontal synchronizing pulses supplied to the synchronizing input 13 and has a frequency which corresponds to an integer multiple of the repetition frequency of the horizontal synchronizing pulses. The frequency of the clock signal is therefore connected to the repetition frequency of the horizontal synchronizing pulses and fluctuates with the changes in this repetition frequency, so that, for example, frequency deviations in this repetition frequency or deviations in the phase position of individual horizontal synchronizing pulses, such as occur in particular in an image recorder, in particular one VCR, supplied, analog television signal occur, lead to changes in the frequency of the clock signal.

The digital television signal from the output 4 of the analog-digital converter 1 is also fed to a television signal input 17 of a demodulator 18, which is connected to a first and a second demodulation signal inputs 19 and 20 also a first and a second demodulation signal are fed. The demodulation signals each consist of a sequence of digital samples with a repetition frequency that is equal to the frequency of the clock signal, and represent sinusoidal oscillations with the frequency of the color synchronization signal (burst) contained in the chrominance signal and a mutual phase shift of 90 ⁰ , so that, for example, the first Demodulation signal represents a sine function and the second demodulation signal represents a cosine function. In the demodulator 18, each sample of the digital television signal from the television signal input 17 is multiplied by a sample of the first and the second demodulation signal. The sequences of the products occurring at the clock frequency are output as demodulated digital color difference signals at a first and a second color difference signal output 21 and 22, respectively. They are supplied on the one hand via a color difference low-pass filter 23 with one channel for each color difference signal for separating high-frequency mixed products formed by the demodulator 18, and on the other hand signal inputs 26, 27 of a first phase detector 25. The clock signal is also fed to the first phase detector 25 at a clock signal input 28.

The first phase detector 25 scans the digital color difference signals fed to it at the signal inputs 26 and 27 at the time of the occurrence of the color synchronization signal in the television signal and thus detects that with the first demodulation signal and at the second signal input 27 that with the second demodulation signal via the first signal input 26 demodulated, ie multiplied, color burst signal. For example, in the case of a television signal transmitted according to the PAL standard, each is followed Horizontal synchronizing pulse transmit an oscillation train of the color synchronizing signal, and accordingly the demodulated color synchronizing signals are detected by the first phase detector 25 in this time interval. The first phase detector 25 forms a color phase signal corresponding to the phase relationship of the color synchronization signal with the demodulation signals and outputs this via a color phase signal output 29 to a color phase signal controller 30, which is preferably constructed in the form of a proportional-integral control stage as a digital filter and has its input 31 supplied color phase signal according to its filter characteristic via an output 32 to a first input 33 of a first adder 34. From the output 35 of the first adder 34, the color phase signal is fed as part of an increment to an increment input 36 of a first phase arithmetic unit 37, to which the clock signal is also fed at a clock input 38. At a control signal output 39, the first phase arithmetic unit 37 emits a first control signal from a sequence of values with a sequence frequency corresponding to the frequency of the clock signal and a difference determined by the increment supplied to the increment input 36.

The first control signal is fed to the input 40 of a signal generator 41. For each possible value of the first control signal, a value of the first and the second demodulation signal is selected in the signal generator 41 and output via a first or a second demodulation signal output 42 or 43 and fed to the first or second demodulation signal input 19 or 20 of the demodulator 18 . The signal generator 41 preferably comprises a read-only memory in which the values of the first and second demodulation signals are stored, which are determined by the values of the first control signal Input 40 can be addressed. Demodulator 18, first phase detector 25, color phase signal controller 30, first adder 34 and the combination of first phase arithmetic unit 37 and signal generator 41 forming a digital oscillator together form a digital phase locked loop, the signal frequency - the frequency of the demodulation signals _{- is} determined by the increment at the increment input 36.

The circuit arrangement according to FIG. 1 further comprises a reference signal source 45 controlled by a quartz crystal 44, to which the clock signal is fed from the clock signal output 15 and which, at an output 46 as a reference signal, outputs a rectangular signal with a frequency which is at least almost equal to the frequency of the color synchronizing signal in TV signal is. This rectangular signal consists of a clock frequency sequence of preferably two different signal values, which follow one another in such a way that a signal of the desired frequency is produced. For this purpose, for example, an analog signal with a rectangular shape and the frequency of the rectangular signal is derived from the quartz crystal 44 and its value is sampled with the clock signal. The rectangular signal can be represented with one bit.

Furthermore, a second phase arithmetic unit 47 is provided, which corresponds in structure and mode of operation to the first phase arithmetic unit 37. It accordingly has an increment input 48, a clock input 49 and a control signal output 50.

The second phase arithmetic unit 47 outputs a second control signal at its control signal output 50, which is sent to a control signal input 51 of a second phase detector 52 is fed. The reference signal from the output 46 of the reference signal source 45 is fed to the second phase detector 52 at a reference signal input 53 and compared in frequency and phase with the second control signal. For this purpose, the differences are formed from the values of the reference signal and the associated values of the second control signal, which represent a measure of frequency or phase deviations, and a reference phase signal formed therefrom is emitted at an output 54 of the second phase detector 52 and, on the one hand, via a reference phase signal controller 55 fed as an increment to the increment input 48 of the second phase arithmetic unit 47 and, on the other hand, to a second input 56 of the first adder 34. The reference phase signal controller 55 is preferably constructed similarly to the color phase signal controller 30 as a proportional-integral control stage. The increment which it outputs to the increment input 48 of the second phase arithmetic unit 47 is additively superimposed on the color phase signal in the first adder 34 and is fed together with it as an increment to the first phase arithmetic unit 37. The increment from the second phase detector 52 or from the reference phase signal controller 55 fluctuates inversely in proportion to the fluctuations in the frequency of the clock signal. From this portion of the increment at the increment input 36 of the first phase arithmetic unit 37, the clock signal supplied to the clock input 38 of the first phase arithmetic unit 37 is produced at the control signal output 39 of the assumption that the color phase signal from the color phase signal controller 30 at the first input 33 of the first adder 34 makes no contribution first phase arithmetic unit 37 a first control signal which immediately follows the reference signal from the output 46 of the reference signal source 45. The color phase signal adds a portion to the increment at the increment input 36 which essentially corresponds to the deviation of the frequencies or phases of the reference signal and the color synchronizing signal from the television signal.

The color phase signal controller 30 preferably has a low-pass characteristic with a very low cut-off frequency, so that the associated phase-locked loop is very frequency-stable, i.e. that it only very slowly tracks fluctuations in the frequency of the color synchronizing signal. In contrast, the reference phase signal controller 55 also enables rapid frequency or phase changes in the phase locked loop comprising it. This phase locked loop is thus able to compensate for rapid fluctuations in the frequency or phase of the clock signal, for example in a television signal fed in by a video recorder after a change in the playback head. Such head changes are usually made at the transition between two successive images in the television signal. The disturbances emanating from these head changes are corrected by the circuit arrangement described before the first television picture line shown on a picture display device appears in the television signal.

In addition to the analog-to-digital converter 1, the first phase detector 25 and the phase arithmetic units 37, 47, the clock signal is also used for at least the filter stage 6, the synchronous separation stage 10, the color difference low-pass filter 23, the color phase signal regulator 30 and the second phase detector 52 and the reference phase signal regulator 55 fed to synchronize their functional processes.

The outputs 7 for the digital luminance signal, 11 for the vertical synchronizing pulses, 16 for the line synchronizing pulses and 24 for the color difference signals in the circuit arrangement according to FIG. 1 are connected, for example, to a television picture display device, not shown.

2 shows a somewhat more detailed circuit diagram for the construction of a phase arithmetic unit used in the circuit arrangement according to FIG. 1 using the example of the first phase arithmetic unit 37. The increment is fed from the increment input 36 to a first input 71 of a modulo adder 70, the output 72 of which is connected to a Input 73 of a register 74 is connected, the output 75 of which supplies the first control signal is connected to the control signal output 39 of the first phase arithmetic unit 37 and to a second input 76 of the modulo adder 70. The clock signal from clock input 38 is also fed to the register in such a way that in each period of the clock signal the signal fed from output 72 of modulo adder 70 to input 73 is stored in register 74 and is available at its output 75. In each period of the clock signal, the value stored in register 74 is thus added to the increment from first input 71 in modulo adder 70, i.e. the increments supplied via the increment input 36 are accumulated in time with the clock signal and starting, for example, with the value zero. In the event of an overflow in the modulo adder 70, i.e. if the sum of the values at inputs 71 and 76 exceeds the largest representable value at output 72, only the difference between this sum value and the largest representable value appears there. The phase arithmetic unit 37 then begins to accumulate the increments again at the increment input 36.

As already shown in FIG. 1, the first phase arithmetic unit 37 is followed by the signal generator 41 at the control signal output 39, which together with the phase arithmetic unit 37 forms a digital oscillator.

3 shows an example of an embodiment of the reference phase signal controller 55 as a proportional-integral control stage, which has a proportional branch with a first Multiplier 80 and, in parallel therewith, an integral branch with an accumulation device comprising an adder 81 and a register 82 and a second multiplier 83 connected downstream of this accumulation device. The clock signal is fed to the register 82 at a clock input 84. The multipliers 80, 83 each have an input 85, 86, via which the reference phase signal controller 55 is supplied with proportionality factors for setting the characteristic, ie the control gain and the time constant and thus ultimately the control speed. These proportionality factors can be fixed, but can also be adjustable. The signals from the proportional branch and the integral branch are summed via a further adder 87.

FIG. 4 shows, as a further embodiment of the circuit arrangement according to the invention, a modification of the arrangement according to FIG. 1, parts which are identical to this circuit arrangement being provided with the same reference numerals and being described with reference to FIG. 1.

In the circuit arrangement according to FIG. 4, a frequency control signal is derived from the clock signal generator 14 via a further output 90 and fed to a first input 92 of a second adder 93 via a conversion stage 91. For example, the clock signal generator 14 comprises a clock signal phase-locked loop with a clock oscillator which outputs the clock signal, a frequency divider which divides the frequency of the clock signal by the frequency of a line synchronizing signal contained in the television signal, and a phase detector which compares the frequency-divided clock signal with the line synchronizing signal with regard to frequency and phase , which forms the frequency signal and thus controls the clock oscillator. The frequency control signal contains immediate information about the frequency of the clock signal.

The second adder 93 is inserted into the signal path for the reference phase signal such that the reference phase signal is fed to its second input 94 and the sum of the reference phase signal and the frequency control signal from its output 95 to the increment input 48 of the second phase calculator 47 and the second input 56 of the first Adder 34 is fed. The conversion stage 91 transforms the frequency control signal, for example, in its value in such a way that it is adapted to the values of the reference phase signal and the color phase signal. In addition, the conversion stage 91 can also perform a dynamic adaptation, for example low-pass filtering or a delay, of the signal determining the frequency of the clock signal. In this way, the mode of operation of the rough regulation of fluctuations in the clock signal carried out by this signal can be adjusted over a wide range. For example, the signal transmission properties in the clock signal generator 14 between the synchronization input 13 and the clock signal output 15 can thus be taken into account.

In a further modification of the circuit arrangement according to FIG. 1, the circuit arrangement according to FIG. 4 contains a reference signal source 100 which emits a digital signal which is periodic at the frequency of the reference signal at an output 101. This is supplied, for example in the form of a multi-digit dual word, to a reference signal input 102 of a second phase detector 103, which differs from the second phase detector 52 of the circuit arrangement according to FIG. 1 in that it compares two signals representing two-digit dual numbers with one another, while the second phase detector 52 from the circuit arrangement according to FIG. 1 the value of the amplitude-discrete second control signal at the time of occurrence of an edge in the rectangular reference signal detected. The amplitude-discrete reference signal supplied by the reference signal source 100 thus has a higher resolution than the rectangular reference signal emitted by the reference signal source 45. An increase in the measuring accuracy of the second phase detector 103 is thus achieved compared to the design of the second phase detector 52 from the circuit arrangement according to FIG. 1. An increase in the control speed of the phase control loop comprising the second phase detector and the second phase arithmetic unit 47 can thus be achieved.

A higher resolution of the reference signal can be achieved, for example, in that the oscillation derived from the quartz crystal 44 is fed to the second phase detector 103 via an analog-digital converter with a high sampling rate compared to the frequency of the reference signal. Another embodiment with a reduced circuit complexity is shown in FIG. 5. An oscillator 130 controlled by the quartz crystal 44 oscillates at a frequency which is at least almost four times the frequency of the color synchronizing signal. A signal with this frequency is output at the output 131 of the oscillator 130 and, on the one hand, is fed directly to a first input 132, via a first divider 133 to a second input 134 and via a second divider 135 to a third input 136 of a three-part register 137. The signal output at the output 131 is preferably rectangular, and each of the three parts of the three-part register 137 assigned to one of the inputs 132, 134, 136 stores one bit. The three-part register 137 is also supplied with the clock signal at a clock input 138. In each period of the clock signal, an instantaneous value of the signals at the inputs 132, 134 and 136 is stored in the three-part register 137 and at the output 139 in the form of a three-digit dual word. This is multiplied in a multiplier arrangement 140 by a predetermined factor and then passed via the output 101 of the reference signal source 100 to the reference signal input 102 of the second phase detector 103.

The multiplier arrangement 140 comprises, for example, a multiplier 141, to which, in addition to the signal from the output 139 of the three-part register 137, a constant multiplication factor is fed from a memory 142 and which multiplies the two together.

A particularly simple embodiment of the multiplication arrangement 140 is obtained if the multiplication factor is an integer power of the decimal number 2. Multiplication by such a multiplication factor merely means shifting the digits of a dual number by a number corresponding to the exponent of the power of two. The multiplier 141 then consists only of adding a corresponding number of lines for the desired number of digits, while the memory 142 simplifies the connection of these lines to ground potential, for example. Such a multiplier arrangement 140 is shown in FIG. 6.

The second phase detector 103 according to FIG. 5 comprises a subtraction stage 143, which receives the reference signal from the reference signal input 102 on the one hand and the second control signal from the control signal input 51 on the one hand and which carries out a modulo subtraction, ie a subtraction without taking into account an occurring carry. If, for example, the multiplication factor of the multiplier arrangement 140 is set to the value 32, the reference signal at the reference signal input 102 has the form of an eight-digit number Dual number. The second control signal is then also supplied in the form of an eight-digit binary number, and the difference is also output by the subtracting stage 143 at its output 144 in the form of an eight-digit dual number. Subtractor 143 then performs modulo (+ 128) subtraction.

From the output 144 of the subtracting stage 143, the signal emitted there is fed to the output 54 of the second phase detector 103 via a recursive filter consisting of an adder 145, a register 146 and a multiplier arrangement 149 formed from a multiplier 147 and a memory 148. Register 146 is advanced by the clock signal supplied via a clock input 150. In the present example, the signal from the output of the register 146 in the multiplier arrangement 149 is multiplied by the factor 7/8. The recursive filter 145 to 149 serves to reduce the so-called quantization noise, i.e. to increase the resolution or to reduce jumps of successive values in the processed signals. The bandwidth of the recursive filter 145 to 149 must be chosen so large that the control speed of the phase-locked loop containing the second phase detector 103 is not noticeably influenced by it.

From the output 54 of the second phase detector 103, the reference phase signal in the circuit arrangement according to FIG. 4 is fed to an input 110 of a distortion stage 111. The distortion stage 111 comprises, for example, a read-only memory with individual memory locations in which values of a reference phase signal to be output at the output 112 of the distortion stage 111 and which is distorted relative to the signal supplied at the input 110 are stored. The individual memory locations are determined by the values of the signal addressed at input 110 and the value of the addressed memory location is supplied to output 112. This allows any, for example quadratic, assignment between the values at input 110 and the values at output 112.

Another example of an embodiment of a distortion stage 111 is shown in FIG. 7. In this circuit, the reference phase signal from the input 110 is supplied on the one hand to an input 160 of a quantizing stage 161 and on the other hand to a reference phase signal input 162 of a selection circuit 163. In the quantizing stage 161, the reference phase signal is compared with a number of threshold values, by means of which the entire range of the values possible for the reference phase signal is divided into a corresponding number of value ranges. The quantizing stage 161 emits a signal at an output 164, which indicates the range of values in which the value of the reference signal that is currently supplied via the input 110 to the distortion stage 111 falls. This signal from the output 164 of the quantizing stage 161 controls two memory stages 165 and 166 contained in the selection circuit 163 in such a way that they each output a signal value at outputs 167 and 168, respectively. The selection circuit 163 further comprises a multiplier 169 and an adder 170. The multiplier 169 multiplies the reference phase signal fed to it from the reference phase signal input 162 of the selection circuit 163 by the signal value supplied by the output 167 of the first memory stage 165, which thus determines the slope of a characteristic curve which determines the assignment between the values of the reference phase signal at input 110 and those at output 112 of distortion stage 111 in the respective value range. That generated by multiplier 169, the product of the reference phase signal from input 110 and the signal value from the first memory Signal representing stage 165 is added in the following adder 170 to the signal value from the second memory stage 166 and supplied as an assigned value of the reference phase signal to the output 112 of the distortion stage 111. With the described distortion stage 111, it is thus possible to make an assignment between the values at the input 110 and at the output 112 in accordance with a continuous, regionally linear characteristic curve. Of course, discontinuous characteristics can also be realized with the arrangement described.

With a distortion stage 111 of the type described, a progressive characteristic curve is preferably set between the values of the reference signal at the input 110 and at the output 112, since in the case of low values for the reference phase signal, a low control gain and thus a low bandwidth and in the case of high bandwidth Values for the reference phase signal causes a high control gain. This enables a more precise correction of phase errors with a low control speed in the case of small phase deviations and a rapid, approximate correction with a high control speed in the case of large phase errors.

The three arrangements described with reference to FIG. 4 for increasing the control speed, namely on the one hand the supply of a signal determining the frequency of the clock signal via the second adder 93, on the other hand the reference signal source 100 delivering an amplitude-discrete reference signal and the third the distortion stage 111 can also be used provide independently of one another and not only together, as shown in FIG. 4, in the circuit arrangement according to FIG. 1.

The circuit arrangement according to FIG. 4 also has a two-stage reference phase signal controller 120, which consists of a first and a second proportional-integral control stage 121, 122 connected in a chain. Each of the stages 121, 122, which are otherwise independent of one another, is constructed, for example, according to FIG. 3.

Instead of the two-stage reference phase signal regulator 120 in FIG. 4, a reference phase signal regulator 180, as shown in FIG. 8, is advantageously also used. It comprises two integrators 181, 182, which are constructed in a manner known per se from an adder 183, 184 and a register 185, 186 in the manner of accumulation devices. For this purpose, the clock signal is fed to the registers 185, 186 at clock inputs 187, 188. Each of the integrators 181, 182 is connected at its output to a multiplier 189, 190, in which each output signal of the integrators 181, 182 is multiplied by a proportionality factor supplied by inputs 191, 192.

The reference phase signal fed from the output 112 to the distortion stage 111 is now, on the one hand, successively via the first integrator 181, the first multiplier 189, the second integrator 182 and the second multiplier 190 to a first input 193 of an output adder 194 and also via a third multiplier 195 to a second one Input 196 of the output adder 194 fed. A further proportionality factor is fed to the third multiplier via an input 197, by which the signal from the output 112 of the distortion stage 111 is multiplied before it is fed to the second input 196 of the output adder 194. In addition, the output adder 194 at a third input 198 receives the signal from the output of the first multiplier 189 forwarded. The sum of the signals from the inputs 193, 196, 198 of the output adder 194 is passed on to the second input 94 of the second adder 93.

Claims (12)

1. Circuit arrangement for deriving digital color signals from an analog television signal with - An analog-to-digital converter (1) which forms a digital television signal from the analog television signal, which comprises a sequence of amplitude-discrete samples of the analog television signal with a repetition frequency determined by a clock signal, which corresponds to an integral multiple of the line frequency of the television signal, a demodulator (18) which forms the digital color signals by multiplying the digital television signal by at least one digital demodulation signal, - a first phase detector (25), the arbsignalen from certain digital _F derives a color phase signal relative to each other a measure of the phase positions of the digital color signals and / or to the demodulation signals, a first phase arithmetic unit (37) which emits a first control signal which consists of a sequence of values with a sequence frequency determined by the clock signal and in which the difference between two successive values is set by the color phase signal, a reference circuit (45, 47, 52, 55) which comprises a reference signal source (45), a phase detector (52) generating a reference phase signal and a second phase arithmetic unit (47) which emits a second control signal consisting of a sequence of signal values, the difference between successive signal values being set by the reference phase signal, and that to the first phase arithmetic unit (37) Additional setting of the difference between two successive values of the first control signal feeds the reference phase signal, characterized in that the sequence of signal values of the second control signal of the second phase arithmetic unit (47) consists of a sequence frequency determined by the clock signal and that the second phase detector (52) also carries the second control signal compares the reference signal generated by the reference signal source (45) and forms the reference phase signal therefrom. 2. Circuit arrangement according to claim 1, characterized in that the reference phase signal is superimposed on the color phase signal before being fed to the first phase arithmetic unit (37). 3. Circuit arrangement according to one of the preceding claims, wherein the clock signal in a clock signal phase-locked loop (14) is derived from a clock oscillator whose oscillation frequency is determined by a frequency control signal obtained from the comparison of a line synchronizing signal contained in the analog television signal with the frequency-divided clock signal characterized in that the frequency control signal is additively superimposed on the reference phase signal supplied to at least the first phase arithmetic unit (37). 4. Circuit arrangement according to one of the preceding claims, characterized in that the reference phase signal to the first and the second phase arithmetic unit (37 and 47) is fed via a reference phase signal controller (55). 5. Circuit arrangement according to claim 4, characterized in that the reference phase signal controller (55) comprises a proportional-integral control stage (Fig. 3). 6. Circuit arrangement according to claim 4, characterized in that the reference phase signal controller (180) comprises a chain connection of at least two integrators (181, 182), which at least one proportional branch is connected in parallel. 7. Circuit arrangement according to claim 4, 5 or 6, characterized in that the reference phase signal an input (110) · a distortion stage (111) and its output signal (112) is fed to the reference phase signal controller (55, 120) and that the distortion stage (111) the reference phase signal is converted such that large values of the reference phase signal are disproportionately enlarged compared to small values according to a non-linear assignment between values at the input (110) and output (112). 8. Circuit arrangement according to claim 7, characterized in that the distortion stage (111) comprises a memory arrangement which for each value of the reference phase signal at the input (110) forms a value at the output (112) of the distortion stage (111) and this the reference phase signal controller (55 , 120). 9. Circuit arrangement according to claim 7, characterized in that the distortion stage (111) comprises a quantizing stage (161) which compares values of the reference phase signal with one or more threshold values, and a selection circuit (163) controlled by the quantizing stage (161), which Inputs (162) converted values of the reference phase signal into individual value ranges limited by the threshold value or values according to a linear assignment into values which are supplied from the output (112) of the distortion stage (111) to the reference phase signal controller (55, 120). 10. Circuit arrangement according to one of the preceding claims, characterized in that the reference signal source (45) emits a rectangular signal which is scanned with a register controlled by the clock signal. 11. Circuit arrangement according to one of the preceding claims 1 to 9, characterized in that the reference signal source (100) comprises an oscillator (130) which emits an oscillation at a frequency which corresponds to an integer multiple of the frequency of the reference signal, and a conversion circuit ( 133, 135, 137, 140), which converts the oscillation of the oscillator (130) into an amplitude-discrete signal that is periodic with the frequency of the reference signal. 12. Circuit arrangement according to one of the preceding claims, characterized in that the frequency of the reference signal is at least almost equal to the frequency of the color synchronizing signal of the television signal.

Images